Question about Training Code

[lh-zhu]

Greetings!

I wish to express my appreciation for your excellent work. I am highly interested in it. I am curious as to whether you could kindly share the training code for the purpose of reproduction. Thank you immensely. I look forward to receiving your early reply.

[MaureenZOU]

I believe this is tested under the code base, I will shortly add it to the read me, thanks!
CUDA_VISIBLE_DEVICES=0 python entry.py train
--conf_files configs/grin/focalt_llama_lang.yaml
--overrides
MODEL.DECODER.HIDDEN_DIM 512
MODEL.ENCODER.CONVS_DIM 512
MODEL.ENCODER.MASK_DIM 512
MODEL.DECODER.NUM_OBJECT_QUERIES 101
FP16 True
MODEL.DECODER.MASK.ENABLED True
MODEL.DECODER.CAPTION.ENABLED True
MODEL.DECODER.SPATIAL.ENABLED True
MODEL.DECODER.RETRIEVAL.ENABLED True
MODEL.DECODER.GROUNDING.ENABLED True
MODEL.DECODER.INTERLEAVE.ENABLED True
MODEL.DECODER.INTERLEAVE.VISUAL_PROB 0.5
SOLVER.MAX_NUM_EPOCHS 50
SOLVER.BASE_LR 0.0001
SOLVER.FIX_PARAM.backbone True
SOLVER.FIX_PARAM.lang_encoder True
SOLVER.FIX_PARAM.pixel_decoder True
COCO.TEST.BATCH_SIZE_TOTAL 1
COCO.TRAIN.BATCH_SIZE_TOTAL 2
COCO.TRAIN.BATCH_SIZE_PER_GPU 2
VLP.TEST.BATCH_SIZE_TOTAL 1
VLP.INPUT.SHORTEST_EDGE True
VLP.INPUT.MIN_SIZE_TEST 512
VLP.INPUT.MAX_SIZE_TEST 864
ADE20K.TEST.BATCH_SIZE_TOTAL 1
REF.TEST.BATCH_SIZE_TOTAL 1
VOC.TEST.BATCH_SIZE_TOTAL 1
FIND_UNUSED_PARAMETERS True
COCO.INPUT.IMAGE_SIZE 640
COCO.INPUT.MIN_SIZE_TRAIN 512
COCO.INPUT.MAX_SIZE_TRAIN 864
COCO.INPUT.MIN_SIZE_TEST 512
COCO.INPUT.MAX_SIZE_TEST 864
VOC.INPUT.MIN_SIZE_TEST 512
VOC.INPUT.MAX_SIZE_TEST 864
MODEL.DECODER.SPATIAL.MAX_ITER 0
MODEL.DECODER.TOP_GROUNDING_LAYERS 6
MODEL.DECODER.TOP_SPATIAL_LAYERS 6
MODEL.DECODER.TOP_RETRIEVAL_LAYERS 6
MODEL.DECODER.TOP_INTERLEAVE_LAYERS 6
MODEL.TEXT.LOAD_TENSOR False
WEIGHT True
FAKE_UPDATE False
RESUME_FROM /disk1/checkpoint/vlcore/xdecoder/vision_xdecoder_focalt_unicl_fb_lang_grin_v1/default/model_state_dict.pt
# MODEL.TEXT.PRECOMPUTE True
# COCO.INPUT.DATASET_MAPPER_NAME coco_interleave_language
# RESUME_FROM /nobackup3/xueyan-data/grin_data/output/xdecoder/vision_xdecoder_davitd3_unicl_fb_lang_grin_v1/default/model_state_dict.pt
# RESUME_FROM /nobackup3/xueyan-data/grin_data/output/xdecoder/vision_xdecoder_davitd5_unicl_fb_lang_grin_v1/default/model_state_dict.pt